KR20170103424A - Apparatus and Method for Collision Detection for Collaborative Robot - Google Patents

Abstract

The present invention relates to an apparatus and a method of detecting a collision of a collaborative robot. According to the present invention, the apparatus to detect a collision comprises: an instruction input unit which receives an instruction position, an instruction speed, and an instruction acceleration to control a motor from the outside, receives information about a current position and a current speed from the motor, and an instruction value input unit which calculates a position error based on the instruction position and the current position; a PID control unit which receives information about the instruction position, the instruction speed, the instruction acceleration, the current position, and the current speed from the instruction value input unit, and generating a proportional integral derivative (PID) control signal based on at least one of them; a threshold value calculation unit which calculates a collision instruction value based on at least one of the instruction position, the instruction speed, and the instruction acceleration; a collision determination unit which determines whether collision occurs based on the position error and the collision threshold value; and the motor operated by the PID control signal, transmitting information about the current position and speed to the instruction value input unit.

Description

[0001] Apparatus and Method for Collision Detection for Collaborative Robot [

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an industrial robot, and more particularly, to an apparatus and method for detecting a collision of a cooperative robot.

In general, the robot industry is evolving day by day, and it is expected to become one of the important industries that determine national competitiveness in the near future. As a result, interest in robotics for educational purposes and toys has been increasing, including home robots for popularization of the robot industry and robot technology. Currently, many companies are increasingly interested in industrial robots as well as home, educational, and toy robots. Currently, many companies are contributing to robot industrialization not only for industrial robots but also for home, education, and toy robots, and it is expected that investment and technology development for intelligent robots will accelerate in the future.

In general, most industrial robots are used for high-load handling and use high-power actuators. Among them, the collaborative robot is required to be closely related to the operator as shown in Fig. 1, so that the safety of the operator is required. Thus, ISO 10218 provides for the safety of robots. The revision of ISO 10218 is currently underway, and new provisions have been added to the cooperation of robots and humans, as well as regulations on the cooperation of multiple robots. ISO 10218 is divided into two parts: Part 1 (Part 1) and Part 2 (Part 2). Part 1 mainly stipulates the safety of the robot itself, that is, the safety that the robot manufacturer should observe. For example, the flange speed should not exceed 0.25ms ^-1 , the maximum power shall be 80W or less, and the maximum static load shall be 150N or less. Part 2 stipulates the stability with which the robot is integrated into the system, that is, the safety that the robot manufacturer, integrator, and user must comply with.

Especially, since collaborative robots work close to each other by robots and workers, safety of collaborative robots is increasingly important. More robust collision detection technology is needed to enhance stability of collaborative robots.

SUMMARY OF THE INVENTION It is an object of the present invention to provide an apparatus and method for detecting a collision of a cooperative robot with enhanced safety.

It is another object of the present invention to provide an apparatus and method for detecting a collision of a cooperative robot with improved reliability of collision detection.

It is another object of the present invention to provide an apparatus and method for detecting a collision of a cooperative robot having a simple operation and a reduced manufacturing cost.

According to one aspect of the present invention, there is provided a collision sensing apparatus for a collaboration robot. The collision sensing apparatus receives information on a command position, command speed and command acceleration for controlling the motor from the outside, receives information about the current position and the current speed from the motor, and outputs the command position and the current position Based on at least one of the command position, the command speed, the command acceleration, the current position, and the current speed from the command value input unit, calculates a proportional integral (PID) a PID controller for generating a derivative control signal, a threshold value calculator for calculating a collision threshold value based on at least one of the command position, the command speed, and the command acceleration, A collision judging unit for judging whether or not a collision has occurred, Value and is implemented by including the motor for transmitting information regarding the current speed portion the command value input.

According to another aspect of the present invention, there is further provided a feedback control unit for receiving a PID control signal from the PID control unit, receiving information about whether a collision has occurred from the collision determination unit, and controlling the motor based on the information.

According to another aspect of the present invention, there is further provided a low-pass filter for obtaining a position error value from the command value input unit and blocking a frequency component higher than a preset cut-off frequency.

According to another aspect of the present invention, there is further provided an encoder for encoding information about the current position and the current speed from the motor and transmitting the encoded information to the command value input unit.

According to still another aspect of the present invention, there is provided a collision detection method performed by a collision sensing apparatus of a collaboration robot. The collision detection method includes: acquiring a command position, an instruction speed, a command acceleration, a current position and a current speed of a motor; calculating a position error based on the command position and the current position; Generating a proportional integral derivative (PID) control signal based on the command position, the command velocity, and the command acceleration, calculating the command position, the command velocity, the command acceleration, Calculating a collision threshold based on any one or more of the weight and the collision threshold, determining whether a collision has occurred based on the command position and the collision threshold, determining whether the collision has occurred, Controlling the motor by generating a motor control signal based on the current position and the current position of the motor, It is implemented, including the step of feeding back the speed.

According to the present invention, there is provided a collision sensing apparatus and method for a collaborative robot in which reliability of collision detection is improved.

An apparatus and method for detecting a collision of a cooperative robot having a simple operation and a reduced manufacturing cost are provided.

1 shows an example in which a collaborative robot performs work with an operator.
2 (a) shows an example of a method of setting a path of a motor using a second-order polynomial model.
2 (b) shows an example of a method of setting a path of a motor using a third-order polynomial model.
FIG. 3 shows an apparatus for detecting a collision of a cooperative robot according to an embodiment of the present invention.
FIG. 4 shows an example of position, velocity, acceleration, position error, and velocity error measured at an arbitrary joint in which no collision occurred.
5 shows an example of position, velocity, acceleration, position error, and velocity error measured at any joint where a crash occurred.
Fig. 6 shows another example of position, velocity, acceleration, position error, and velocity error measured at any joint where a crash occurred.
7 is a block diagram of an apparatus for detecting a collision of a cooperative robot according to an embodiment of the present invention.
FIG. 8 shows an example of a signal extracted by the low-pass filter 740 in FIG.
9 is a flowchart of a method of detecting a collision of a cooperative robot according to an embodiment of the present invention.

Hereinafter, some embodiments will be described in detail with reference to exemplary drawings. It should be noted that, in adding reference numerals to the constituent elements of the drawings, the same constituent elements are denoted by the same reference symbols as possible even if they are shown in different drawings. In the following description of the embodiments of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present disclosure rather unclear.

를 움직이기 위해 요구되는 토크(desired torque)는 아래의 수학식 1에 의하여 계산될 수 있다.Collaborative robots can realize collision detection by using a large variation in parameters of each axis of the robot when a contact with a worker occurs while performing a task. Dynamics can be utilized to realize collision detection of collaborative robots. For example, if a desired desired angle of the robot is desired,

The desired torque to move the torsion spring can be calculated by the following equation (1).

는 관성 행렬(inertial matrix),

는 코리올리/원심력(Coriolis/centrifugal) 행렬(matrix),

는 중력항,

는 마찰항을 나타낸다. 하지만, 실제로 충돌이 발생하는 경우는 충돌에 의한 토크(

)를 고려해야 한다. 충돌에 의한 토크(

)는 실제 모터의 전류로부터

값을 측정하고, 이 값을 이용하여 도출할 수 있다. 상기 충돌에 의한 토크(

)는 아래 수학식 2를 이용하여

값으로부터 계산될 수 있다.In Equation (1)

An inertial matrix,

A Coriolis / centrifugal matrix,

Is a gravitational term,

Represents a friction term. However, when a collision actually occurs, the torque due to the collision

). Torque due to collision (

) From the actual motor current

The value can be measured and derived using this value. The torque due to the collision (

) Is calculated using Equation 2 below

Lt; / RTI >

은 주위환경, 마모 정도 등에 따라 변화하여 정밀한 수식모델을 구성하는 데 한계가 있다. 이 때, 각각의 조인트에 토크센서를 부착하면 복잡한 계산식이 필요 없고 정밀한 토크 측정이 가능하여 활용도가 높다. 다만, 여러 개의 토크센서를 부착하는 데에 따른 로봇의 비용이 증가하는 단점이 있다. In the case of a multi-axis robot having several joints, it is necessary to perform a much more complicated operation than the above Equations (1) and (2). Therefore, there are many difficulties in calculating the torque due to the collision of the multi-axis robot in real time. Further, the frictional resistance of Equation 1

There is a limitation in forming a precise mathematical expression model by changing with the surrounding environment, the degree of wear, and the like. At this time, attaching a torque sensor to each joint eliminates the need for complicated calculation formulas and allows precise torque measurement to be used. However, there is a disadvantage in that the cost of the robot is increased due to attachment of a plurality of torque sensors.

Hereinafter, a collision sensing apparatus and method according to an embodiment of the present invention capable of avoiding complex mathematical models and reducing manufacturing costs will be described.

Fig. 2 shows an example of a path setting method for controlling the robot. 2 (a) shows a method of setting a path of a motor using a second-order polynomial model, and FIG. 2 (b) shows a method of setting a path of a motor using a third-order polynomial model It shows how to set. Referring to FIGS. 2 (a) and 2 (b), the motor can be controlled by setting the position, speed, and acceleration for each time to control the robot to move from each position to the next target point.

3 shows a collision sensing apparatus 300 of a cooperative robot according to an embodiment of the present invention. 3, the collision sensing apparatus 300 includes a command value input unit 310, a tracking error calculation unit 320, a feed forward torque calculation unit 330, a PID (proportional integral derivative) A torque summing unit 350, a current control unit 360, an encoder 370,

The command value input unit 310 receives information on the position, speed, and acceleration for controlling the motor from the outside.

), 지령 속도(

), 지령 가속도(

)에 대한 정보를 수신하고, 인코더(370)로부터 모터의 실제 위치(

), 실제 속도(

)를 수신하여, 지령 위치와 실제 위치를 비교한 값으로부터 위치오차를 계산하고, 지령 속도와 실제 속도를 비교한 값으로부터 속도오차를 계산한다.The tracking error calculator 320 calculates a tracking error value from the command value input unit 310,

), Command speed (

), Command acceleration (

And receives information from the encoder 370 on the actual position of the motor (

), Actual speed (

), Calculates the position error from the value obtained by comparing the command position and the actual position, and calculates the speed error from the value obtained by comparing the command speed and the actual speed.

The feedforward torque calculation unit 330 calculates the feedforward torque value from the position, speed, and acceleration values input through the command value input unit 310 and transmits the calculated feedforward torque value to the torque summing unit 350.

The PID compensator 340 generates a PID control value based on a control value received from the tracking error calculator 320 and a preset proportional integral derivative (PID) gain schedule, Lt; / RTI >

The torque summing unit 350 sums the feedforward torque value received from the feedforward torque calculation unit 330 and the PID control value received from the PID compensator 340 and transmits the sum to the current controller 360.

The current control unit 360 calculates a current value to be applied to the motor for controlling the arm of the robot based on the torque value received from the torque summing unit 350 and transmits the current value to the encoder 370.

The encoder 370 encodes the current value received from the current controller 360 to control the motor provided in the robot arm, measures the actual position of the motor, and feeds back the measured value to the tracking error calculator 320.

According to the configuration of the collision sensing apparatus 300, the position error value between the command position and the actual position input from the outside can be calculated, and it is possible to determine whether a collision has occurred in the robot using the position error value have.

), 속도(

), 가속도(

), 위치오차(

), 속도오차(

)의 일례를 나타낸다. 모터 제어기의 내부에서는 모터에 부가하는 전류를 계산하기 위하여 위치, 속도, 가속도를 이용한다. 모터에 충격이 발생하는 경우에는 상기 세 가지 값이 급격하게 변화한다. 속도 및 가속도의 경우에는 측정된 디지털화된(digitalize)된 위치값을 미분하여 측정하기 때문에 기본적으로 제어기의 사이클(cycle)과 비슷한 고주파수(high frequency)의 변화로 나타나지만, 상기 속도 및 가속도의 측정을 통해서는 충돌이 있는 경우와 없는 경우를 구분하기 어렵다. 하지만, 도 5 및 도 6에 나타난 것과 같이 위치오차를 참조하면 충돌이 있는 경우와 충돌이 없는 경우를 명확하게 구분할 수 있다. 따라서, 위치 오차의 크기가 미리 설정된 충돌 임계값(collision threshold)보다 클 경우에는 충돌이 발생한 것으로 판단하고, 이에 따른 적절한 안전 조치를 취할 수 있다.Fig. 4 is a graphical representation of the position measured at any joint in which no collision occurred in the motor controller according to Fig. 3

), speed(

), Acceleration (

), Position error

), Speed error (

). Inside the motor controller, position, velocity, and acceleration are used to calculate the current added to the motor. When an impact occurs in the motor, the three values change abruptly. In the case of speed and acceleration, since the measured digitalized position value is measured by differentiating it, it appears basically as a high frequency change similar to the cycle of the controller. However, by measuring the speed and acceleration It is difficult to distinguish between the case where there is a collision and the case where there is no collision. However, referring to the position error as shown in FIG. 5 and FIG. 6, the case where there is a collision and the case where there is no collision can be clearly distinguished. Accordingly, when the magnitude of the position error is larger than a predetermined collision threshold, it is determined that a collision has occurred, and appropriate safety measures can be taken accordingly.

), 속도(

), 가속도(

), 위치오차(

), 속도오차(

)의 일례를 나타낸다. 도 5를 참조하면, 시간 t=2.3sec 일 때 충돌이 발생하였고, 이 때 위치오차(

)가 미리 설정된 충돌 임계값인 66을 초과하게 되어 이로부터 충돌이 발생한 것으로 판단될 수 있다.Fig. 5 is a graph showing the relationship between the position measured at any joint at which a collision occurred in the motor controller according to Fig. 3

), speed(

), Acceleration (

), Position error

), Speed error (

). Referring to FIG. 5, a collision occurred at a time t = 2.3 sec. At this time, the position error

) Exceeds 66, which is a preset collision threshold value, from which it can be judged that a collision has occurred.

), 속도(

), 가속도(

), 위치오차(

), 속도오차(

)의 다른 일례를 나타낸다. 도 6을 참조하면, 시간 t=0~6 sec의 구간에서는 위치오차(

)가 상승하는 구간이 발생한다. 특히, t=5.5sec일 때 가속도의 지령값이 크고 위치오차도 일반적인 구간에 비하여 크지만, 위치 오차가 상기 시간에서의 미리 설정된 충돌 임계값인 216보다는 작아 충돌이 발생하지 않은 것으로 판단될 수 있다. 반면에 t=10.5sec에서는 가속도의 지령값과 속도의 지령값은 낮지만 위치 오차가 상기 시간에서의 미리 설정된 충돌 임계값인 21보다 크기 때문에 충돌이 발생한 것으로 판단될 수 있다.FIG. 6 is a graph showing the relationship between the position measured at any joint where a collision occurred in the motor controller according to FIG. 3

), speed(

), Acceleration (

), Position error

), Speed error (

). ≪ / RTI > Referring to FIG. 6, in a period of time t = 0 to 6 sec,

) Is generated. In particular, when t = 5.5 sec, the command value of the acceleration is large and the position error is larger than the normal section, but it can be judged that no collision occurs because the position error is smaller than the predetermined collision threshold value 216 at the time . On the other hand, at t = 10.5 sec, the command value of the acceleration and the command value of the velocity are low, but it can be judged that the collision occurs because the position error is larger than the predetermined collision threshold value 21 in the time.

The size of the position error varies depending on the weight of the workpiece, the position of the path, and the speed of the robot. When a collision occurs, it also varies with the weight of the workpiece, the position of the path, the speed, and the like. In the present invention, the threshold value may be configured according to Equation (3) according to the magnitude of the acceleration command.

In Equation (3), n means the sensitivity value of the collision, and can be generally set to a value of 2-5. If the value of n is set to a small value, the collision threshold value becomes small, so that the controller can be configured to be sensitive to the collision. However, if the value of n is set to a too small value, the collision can be erroneously determined as collision. On the contrary, if the value of n is set to a large value, the collision threshold value becomes large, so that the controller is not sensitive to the collision. If the value of n is set to a too large value, the collision may not be recognized.

7 is a block diagram of a collision sensing apparatus for a cooperative robot according to an embodiment of the present invention.

7, the collision sensing apparatus 700 includes a command value input unit 710, a PID control unit 720, a low pass filter 730, a threshold value calculation unit 740, a collision determination unit 750, A feed back controller 760, a motor 770, a robot arm 780, and an encoder 790.

The command value input unit 710 receives information on command positions, command speeds, and command accelerations for controlling the motor 770 from the outside and receives information about the current position and the current speed of the motor 770 from the encoder 790 . The command value input unit 710 compares the command position of the motor 770 with the current position to calculate a position error, and compares the command speed and the current speed to calculate a speed error.

The PID controller 720 calculates a proportional integral derivative (PID) control signal 720 based on the command position, command speed, command acceleration, current position of the motor input from the encoder 790, And transmits it to the feedback control unit 760. [ The PID control signal may be a current signal to be applied to the motor 770.

The collision threshold calculator 730 calculates the collision threshold based on at least one of the command position, the command speed, the command acceleration, or the weight of the robot arm. The command position, the command speed, and the command acceleration can be received from the command input unit 710, and the weight of the robot arm can be preset in the system. The collision threshold can be calculated by Equation (3). In Equation (3), a predetermined value may be used as the value of n, and may be set based on one or more of the command position, command speed, and the weight of the robot arm input to the command value input unit 710. For example, when the command speed is high, the value of n is set to a small value so that it is more sensitive to collision. When the command speed is low, the value of n is set to a large value so as to be less sensitive to collision Can be set. The collision threshold value may be set to a different value depending on the weight of the robot arm. For example, when the weight of the robot arm is heavy, the value of n is set to a small value, so that it is more sensitive to collision. When the weight of the robot arm is light, the value of n is set to a large value, Or the like. Meanwhile, the collision threshold value may be set to a different value according to the acceleration command value.

The low pass filter 740 passes all frequency components below a predetermined cut off frequency among the position error values received from the command value input unit 710 and outputs a frequency component higher than the cut- . Therefore, only the change of the position error due to the actual collision is extracted from the position error calculated by the command value input unit 710. The low-pass filter 740 derives an output signal from the input signal using Equation (4) below.

In Equation (4), x _i represents an input signal, y _i represents an output signal, t represents a sampling interval, and f _c represents a cutoff frequency. FIG. 8 shows an example of a signal extracted by the low-pass filter 740. In the example of Figure 8 has been set to 1msec = t, f _c = 10Hz, a = 0.6, collision threshold (collision threshold) = 0.06. Referring to FIG. 8, a frequency component higher than a cutoff frequency is blocked by the low-pass filter 740, thereby eliminating an error that erroneously detects a collision if no actual collision occurs. In the example of Fig. 8, since the position error exceeds the collision threshold value in the vicinity of t = 2.5 sec, the collision can be detected as occurring at this time. Depending on the configuration, the low-pass filter 740 may be omitted.

The collision determining unit 750 determines whether a collision has occurred based on the position error extracted by the low-pass filter 740 and the collision threshold calculated by the collision threshold calculator 730. If the absolute value of the position error is larger than the collision threshold value, it is determined that a collision has occurred. If the absolute value of the position error is smaller than or equal to the collision threshold value, it can be determined that collision does not occur. Meanwhile, the collision determining unit 750 may determine whether a collision has occurred by comparing the velocity error with an arbitrary collision threshold value, and may determine that the velocity error can be used as an auxiliary means for determining whether a collision has occurred have. For example, if the absolute value of the position error is greater than the collision threshold and the absolute value of the velocity error is greater than any other collision threshold, or if one of the two conditions is met, It can be judged that it has occurred. The collision determining unit 750 transmits information indicating whether or not a collision has occurred to the feedback control unit 760.

The feedback control unit 760 receives the PID control value from the proportional integral derivative (PID) control unit 720, receives information about the occurrence of the collision from the collision determination unit 750, and controls the motor 770. For example, when no collision has occurred, the feedback control unit 760 controls the motor by applying the current value received from the PID controller 720 to the motor, and when a collision occurs, Can be stopped.

The motor 770 rotates based on the control signal received from the feedback control unit 760 to operate the robot arm 780. On the other hand, the motor 770 transmits information on the current position and the current speed to the encoder 790.

The encoder 790 encodes information on the current position and current speed of the motor received from the motor 770, and transmits the encoded information to the command value input unit 710. The information on the current position and the current speed of the motor transmitted to the command value input unit 710 can be used to calculate the position error. Meanwhile, the encoder 790 may be omitted, and information on the current position and the current speed of the motor 770 may be directly transmitted from the motor 700 to the command value input unit 710.

According to the configuration of the collision sensing apparatus 700, a position error value between an instruction position and an actual position input from the outside can be calculated, and the position error value is compared with the collision threshold value to determine whether a collision has occurred in the robot .

9 is a flowchart of a collision detection method performed in the collision sensing apparatus for a cooperative robot according to an embodiment of the present invention.

Referring to FIG. 9, the collision sensing apparatus 700 acquires information on the command position, command speed, command acceleration, current position of the motor, and current speed for controlling the motor (S910). Information on the command position, the command speed, and the command acceleration can be obtained by an external input, and information on the current position and the current speed of the motor can be obtained from the encoder 790.

Next, the collision sensing apparatus 700 calculates a position error and a velocity error (S920). The erroneous error can be calculated by comparing the command position with the actual position, and the speed error can be calculated by comparing the command speed and the actual speed.

Next, the collision sensing apparatus 700 generates a proportional integral derivative (PID) control signal from the command position, the command speed, and the command acceleration value (S930). The PID control signal may be a current signal to be applied to the motor.

Next, the collision sensing apparatus 700 calculates a collision threshold value based on at least one of the command position, the command speed, the command acceleration, or the weight of the robot arm (S940). The command position, the command speed, and the command acceleration may be acquired by an external input in step S910, and the weight of the robot arm may be preset in the system. The collision threshold can be calculated by Equation (3). The value of n used in Equation (3) may be set to a predetermined value or may be set based on one or more values of the command position, the command speed, and the weight of the robot arm. For example, when the command speed is high, the value of n is set to a small value so that it is more sensitive to collision. When the command speed is low, the value of n is set to a large value so as to be less sensitive to collision Can be set. The collision threshold value may be set to a different value depending on the weight of the robot arm. For example, when the weight of the robot arm is heavy, the value of n is set to a small value, so that it is more sensitive to collision. When the weight of the robot arm is light, the value of n is set to a large value, Or the like. Meanwhile, the collision threshold value may be set to a different value according to the acceleration command value.

Next, the collision sensing apparatus 700 determines whether a collision has occurred based on the position error and the collision threshold (S950). The position error may be a value calculated in step S920, and an output value obtained by passing the position error calculated in step S920 to a low pass filter may be used. The low-pass filter can derive the output signal from the input signal using Equation (4). The collision detection entity 700 determines whether a collision has occurred based on the position error and the collision threshold value calculated in step S940. If the absolute value of the position error is larger than the collision threshold value, it is determined that a collision has occurred. If the absolute value of the position error is smaller than or equal to the collision threshold value, it can be determined that collision does not occur. The result of the determination as to whether the collision has occurred may be provided by generating a series of signals.

Next, in step S960, the collision sensing apparatus 700 generates a control signal based on the PID control signal generated in step S930 and the determination result of whether or not the collision generated in step S950 has occurred. The control signal may be a current value applied to the motor. For example, when a collision does not occur, the PID control signal is transmitted to the motor to control the motor, and when a collision occurs, no signal is transmitted to stop the motor. When a collision occurs, the collision sensing apparatus 700 can terminate all operations for the safety of the operator.

Finally, the collision sensing apparatus 700 operates the robot arm and feeds back information on the current position and current speed of the motor (S970). Information on the current position and the current speed of the motor may be fed back to the command value input unit 710. Information on the current position and the current speed of the motor to be fed back may be used as a value for obtaining a position error in step S920 have.

According to the present invention, the reliability of collision detection is improved, and the computation is simple, so that the cost required to implement the collision detection in the collaboration robot can be reduced.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments. It will be apparent to those skilled in the art that various modifications may be made without departing from the scope of the present invention. Accordingly, the true scope of the present invention should be determined by the technical idea of the appended claims.

Claims (2)

A collision sensing apparatus for a collaborative robot,
And a controller for receiving information on a command position, an instruction speed and a command acceleration for controlling the motor from the outside, receiving information on the current position and the current speed from the motor, and based on the command position and the current position, A command value input unit for calculating the command value;
A PID controller for receiving information on the command position, the command speed, the command acceleration, the current position, and the current speed from the command value input unit and generating a PID (proportional integral derivative) control signal based on at least one of the command position, the command acceleration,
A threshold calculator for calculating a collision threshold value based on at least one of the command position, the command speed, and the command acceleration;
A collision judging unit for judging whether a collision has occurred based on the position error and the collision threshold value;
The motor operating by the PID control signal and transmitting information about a current position and a current speed to the command value input unit;
A feedback control unit for receiving the PID control signal from the PID control unit, receiving information about the occurrence of a collision from the collision determination unit, and controlling the motor based on the received information;
A low-pass filter for obtaining a position error value from the command value input unit and blocking a frequency component higher than a preset cut-off frequency; And
An encoder for encoding information on the current position and the current speed from the motor and transmitting the encoded information to the command value input unit,
And a collision sensing device.
A collision sensing method performed by a collision sensing device of a collaboration robot,
Obtaining command position, command speed, command acceleration, current position and current speed of the motor;
Calculating a position error based on the command position and the current position, and calculating a speed error based on the command speed and the current speed;
Generating a proportional integral derivative (PID) control signal based on the command position, the command velocity, and the command acceleration;
Calculating a collision threshold value based on at least one of the command position, the command speed, the command acceleration, and the weight of the robot arm;
Determining whether a collision has occurred based on the command position and the collision threshold;
Generating a motor control signal based on the PID control signal and whether or not the collision has occurred to control the motor; And
Feeding back the current position and current speed of the motor;
Lt; / RTI >

Images